Hue- and luminance-preserving saturation enhancement in RGB color space based on dark channel prior

In this study, we propose a hue-preserving saturation enhancement method that guarantees processing within the color gamut in the RGB color space and preserves the luminance perceived by humans. In the proposed method, image saturation enhancement is achieved by changing the pixel values on an equi-luminance line within an equi-hue plane consisting of the same hue and luminance as those of the input pixels. Specifically, γ correction is performed on an equi-luminance line to enhance the saturation. This method allows users to freely adjust the vividness of images by changing parameter γ which controls the degree of enhancement. Furthermore, we propose a method for determining the optimal degree of saturation enhancement based on the dark channel prior (DCP). The DCP represents a statistical characteristic of haze-free outdoor images: most local patches in haze-free outdoor images contain pixels with very low intensities in at least one color channel. By reconsidering this prior knowledge from the perspective of saturation, the proposed method enables the determination of effective saturation enhancement levels for each pixel. In the experiment, the effectiveness of the proposed method was verified by qualitatively and quantitatively comparing it with conventional methods using various images. The experimental results verified that the proposed method can effectively enhance the saturation in a natural appearance. In addition, the subjective evaluation showed that the proposed method based on the DCP was evaluated favorably on average and subjectively.